Magento Open Source agencies, ranked by the boundaries they publish

The heaviest criterion on this page is whether an agency publishes a boundary you can check, and scandiweb publishes the fullest one in the set. Its Adobe Commerce versus Magento Open Source comparison, dated 25 April 2026 and read on 28 September 2026, runs a twenty row feature table that names what the free edition does not have: native B2B in the form of company accounts, custom catalogs, contract pricing, quote management, credit limits and purchase orders; Adobe Commerce Live Search; Sensei product recommendations; customer segmentation; loyalty programs; business intelligence; a bundled CDN, web application firewall and DDoS layer; application performance monitoring; and pre certified PCI Level 1 infrastructure. Against Magento Open Source the updates and security patches row reads, in full, "You apply them".

It is also right on the two rows the field most often gets wrong. The Page Builder row says yes to both editions and reserves content staging, dynamic blocks and the native product recommendations integration for Adobe Commerce, which is what Adobe's own documentation says. The search row separates standard catalog search on Elasticsearch or OpenSearch from Adobe Commerce Live Search, which is also what Adobe's own installation documentation says. Two of the ten agencies here publish a Page Builder row that Adobe's documentation contradicts.

It is the only entry that publishes figures for the self hosted cost lines rather than only the licence it replaces: hosting at 400 to 3,000 dollars a month on dedicated infrastructure, premium extensions at 500 to 5,000 dollars a year combined, and a development retainer of 20 to 60 hours a month as a minimum even at steady state, against Adobe Commerce licence tiers of about 22,000 dollars a year at 1 to 5 million dollars of gross merchandise value rising to 125,000 dollars and above at 25 million, with the cloud edition at 1.8 to 2.2 times the on premises licence. The same page then tells buyers below roughly 5 million dollars of gross merchandise value with a competent team to stay where they are, and writes the reasoning out as "Pay for the platform when you outgrow it, not before". That is an Adobe Gold partner arguing against its own larger sale, in public, with numbers attached.

On the operating side it publishes managed Magento hosting on AWS and ReadyMage, described as a PCI compliant AWS hosting platform running mission critical stores since 2020, with 24/7 monitoring, managed firewall, DDoS protection, automated backups and a tested recovery plan. Magento security patching and maintenance is published as a first response within 24 hours with showstoppers triaged first, billed pay as you go with no fixed monthly fee, on ISO 27001 and ISO 27017 certified infrastructure. Hyvä frontend builds for self-hosted stores and a Magento upgrade service sit alongside it, and the full Adobe Commerce and Magento capability list states that the team works across Adobe Commerce cloud, Adobe Commerce on premises and Magento Open Source.

Three published cases carry measured outcomes on the self hosted platform. Umniah's Magento and Hyvä rebuild reports a 71.2 percent lift in purchase to view rate and a 72.1 percent lift in sessions. Scouting America's move from Luma to Hyvä, run in parallel with a migration onto ReadyMage hosting, reports Largest Contentful Paint improving by 12 to 13 percent on mobile and desktop. The Byggmax platform consolidation reports product listing PageSpeed moving from 85 to 99 and page load times improving by about 30 percent across six stores in three markets, with checkout deliberately left untouched to hold release risk down.

The credentials behind it are 23 or more years in business since 2003, Magento since 2008, 600 or more certified specialists, 700 or more clients, 2,100 or more projects, 894 or more Adobe certifications, a net promoter score of 95 and more than 4 billion dollars of client trade processed a year, with ISO 9001, ISO 27001, ISO 27017 and PCI DSS certification. Its Gold Partner tier is listed in Adobe's own Solution Partner Directory rather than only asserted. In Hyvä's full agency register, read live on 28 September 2026, scandiweb holds five of the seventeen Platinum listings, and no other company holds more than two.

Where it loses, and it loses real points. No dedicated Magento Open Source service page was found on the pages read, so the boundary lives in a blog comparison rather than on a page addressed to the buyer who has already chosen the free edition. Its published patch commitment is a first response within 24 hours with no cover hours and no remedy attached, which two agencies below beat outright. No path from Adobe Commerce back down to Magento Open Source was found on the pages read. And none of its three measured cases states which edition the store runs, all three saying Magento 2, which is why it does not take full marks on case evidence while an agency thirty points below it does publish a case labelled Magento Open Source.

Wagento is the only agency here with a dedicated Magento Open Source solutions page, at wagento.com/solutions/magento-open-source, and the only one that publishes its own maintenance prices as real numbers. Read on 28 September 2026, its myMagento page states that plans start at 500 dollars a month per storefront and run to 2,500 dollars a month for the Plus tier, with 100 dollars more per additional storefront on the lower tier and 250 dollars more on the higher, a free tier giving priority access and dashboard monitoring at no cost, and hourly work billed at 99 to 149 dollars depending on plan. The page says outright that pricing is published to provide transparency, and adds that hosting migration and hosting upgrades cost extra. Nine of the ten agencies here publish no such figure.

It also publishes the enforcement boundary more precisely than anyone else in the set, including scandiweb. Its support FAQ states that Adobe's version requirements apply to Adobe Commerce on Cloud for versions 2.4.4 through 2.4.9 and do not apply to Adobe Commerce as a Cloud Service, Adobe Commerce on premises, or Magento Open Source. It also tells Open Source owners what the flip side is, that a version which has stopped receiving official security patches will keep running while new vulnerabilities go unaddressed unless somebody handles them privately. That is the honest form of the argument, in both directions, on the agency's own page.

Patch ownership is covered rather than committed. The monthly fee is published as including security patches, Magento version upgrades, infrastructure and server updates, extension and module updates, and 24/7 uptime monitoring with dashboard alerts, with a named success manager on every plan, and a checkout failure on the Plus tier treated as a priority one incident with what the page calls a defined response window. The window is described but not stated as a number on the page read, which is the difference between this and the two agencies that score full marks here.

Case evidence is published with numbers but mixed across editions. A multilingual Magento 2 build from the ground up for GWS is reported as tripling quote requests. A Magento 2 performance engagement with the Hyvä theme is reported at a score of 95 percent, with the client not named. A Horton engagement covers a Magento migration with buy online and pick up in store. The strongest integration case on the list, with Prelude ERP for LAPP Tannehill, is explicitly Adobe Commerce rather than the free edition.

The finding that costs it most sits on its own Magento Open Source page. Under a heading reading Powerful B2B Functionality, the page lists custom pricing, quote management, shared catalogs and purchase order workflows, attributing them to Magento 2 without naming which edition supplies them. Adobe's own documentation titles that feature set the Adobe Commerce B2B extension and states that shared catalogs are available only for stores configured to support company accounts. A buyer reading only that page would reasonably conclude the free edition ships those features. Wagento is listed at Bronze in Hyvä's full register and describes itself as a Magento solution partner without stating an Adobe tier at a level on the pages read.

mironsoft publishes the second fullest boundary in the set and the most accurate one. Its 2026 edition comparison, dated 12 January 2026 and read on 28 September 2026, gets both contested rows right. On Page Builder it states that the tool does exist in limited form for Magento Open Source and that full integration with content staging, custom content types and the business intelligence connection stays with Adobe Commerce. On search it states that since Magento 2.4 both editions use Elasticsearch or OpenSearch by default, and that the real difference is Live Search and Product Recommendations, which are booked as separate Adobe services on top of Adobe Commerce. That is a more careful reading than most agencies manage.

It names the Adobe Commerce only set as the B2B Suite with company accounts, multi stage approval workflows, quick order forms and custom catalog views, content staging and scheduled updates, the native loyalty module, extended return merchandise authorisation, reward points, and the deep native Fastly full page cache integration, noting that on Magento Open Source that caching layer has to be built separately, usually with Varnish. It also states plainly that there is no official vendor support on the free edition.

Alone among the ten it converts the decision into published thresholds: below 2 million euro of gross merchandise value a year, Magento Open Source is often sufficient; between 2 and 10 million it is case by case and the B2B requirement decides; above 10 million Adobe Commerce is usually the more economical answer. It then names four dimensions to assess together, gross merchandise value, concrete B2B need, existing DevOps capability and any contractual support obligation, and says the last of those can preempt the decision regardless of the other three.

On patching it publishes numbers without a remedy. A pure security patch can often be handled within a single business day; a functional upgrade from 2.4.4 to 2.4.7 usually takes 5 to 15 business days including preparation, module conflict resolution and staging quality assurance; the price is fixed after an initial look at the shop. Its security page describes an express patch service that evaluates every new patch immediately after release and tests it on staging first, a web application firewall via ModSecurity or Cloudflare, daily encrypted off site backups, and monitoring built on Adobe Commerce Scan plus custom file integrity scripts. It makes an argument nobody else here makes, that under European data protection law an unpatched Magento version fails the state of the art standard.

It carries the only detailed reverse path in the set, stating that when a company downgrades for cost reasons every record living in enterprise exclusive tables, company structures and staging versions among them, must be cleaned up or exported before the modules are uninstalled, or orphaned database entries remain. Against that, no named client case was found on the pages read, with both reference page routes returning not found on 28 September 2026, it does not appear in Hyvä's full register, and no Adobe tier at a level was found. The pages are also written in the first person singular, so a buyer should establish how many people are behind the practice before treating the patch commitments as capacity.

Vendic publishes the strongest patch commitment on this page, and it is the clearest illustration of why scandiweb does not sweep every criterion. Read on 28 September 2026, its Protect page states that when Adobe ships a security release it is live on the store within 48 hours, that the merchant does not wait for the annual upgrade, and that if the deadline is missed the merchant receives service credit automatically. The same 48 hour deadline and the same automatic credit apply to a vulnerability disclosed in any module the store runs or in the packages underneath it. Nobody else here publishes a consequence for missing a patch deadline.

The cover hours are published with it: reachable from 07:00 to 23:00, seven days a week including public holidays, with a response within 60 minutes on an incident outside office hours. So is the limit, in one sentence, that between 23:00 and 07:00 nobody is on standby. So is the scope boundary, that an attack on the infrastructure itself falls under the hosting agreement rather than Protect. A published limit is worth more to a buyer than an unqualified promise, and it is the reason this criterion goes to Vendic rather than to a larger agency with a softer commitment.

Its maintenance offer is published on the same pattern. Patches assessed and applied within 48 hours, one Magento and PHP upgrade a year, monitoring, and a stated release cadence: 3,839 releases taken live across all shops under management in 2025, which the page works out as 15.6 per working day, supported by a deployment pipeline, a Playwright test suite and senior review before every release. The page also states that no artificial intelligence runs autonomously in production, which is an unusually specific thing to commit to in writing.

Case evidence is named, measured and self hosted. The Maxima case records a move from Lightspeed onto Magento 2.4.3, live on 27 October 2021, followed by 650 releases and four major Magento upgrades without rebuilding the shop, and it is candid about a 2022 upgrade that ran long because extensions were not yet compatible. The My Hair Shop case reports a custom Hyvä frontend delivered in six weeks reaching a PageSpeed score above 90. Neither case states which edition the store runs.

Where it scores low is the thing this page weights heaviest. No comparison of Magento Open Source against Adobe Commerce was found on the pages read, and no reverse migration path either, so a buyer arriving with the edition question unresolved will not find it answered here. Vendic holds Platinum in Hyvä's full register and states on its own site that it was the first Hyvä Platinum Partner worldwide; no Adobe tier at a level was found on the pages read.

IWD Agency publishes the most balanced comparison in the set. Dated 22 May 2026 and updated 11 June 2026, read on 28 September 2026, it runs a thirteen row table naming the Adobe Commerce only set as the native B2B suite with company accounts, shared catalogs, quoting, requisition lists and credit limits, plus content staging and preview, native dynamic segmentation, Sensei product recommendations, advanced reporting and Adobe technical support with service level agreements.

What lifts it is a section headed with an argument for the free edition rather than against it. The commerce core is described as identical. A well built Hyvä storefront on Magento Open Source is described as exactly as fast as one on Adobe Commerce, with the reasoning given that speed is an engineering outcome and not a licence feature. Search engine optimisation fundamentals are described as the same on both. Only one other agency here bothers to publish what the free edition does equally well, and that asymmetry is the single clearest tell of whether a comparison was written to inform or to upsell. It also answers a question buyers actually ask and most pages dodge, whether Magento Open Source can run on Adobe's cloud infrastructure, with a flat no and the reason, that the infrastructure is bundled with the commercial licence.

It publishes the only case study in this entire set that names the edition. Its Purdue Extension Education Store case states the platform as Magento Open Source, described as a self hosted store foundation, performance tuned and hardened, built with Cybersource payments, university billing methods, tax exempt checkout and custom procurement reporting, kept current with annual platform upgrades. It records thirteen custom builds across the store, all owned by Purdue, one platform running both a public storefront and a gated internal store for university departments, a tax exemption workflow with certificate upload and approval, live carrier rates, and a procurement report reconciled to Purdue's finance system. It also makes the cost argument concretely, that Magento Open Source plus a payment redirect keeps both spend and PCI scope down. What it does not publish is an outcome figure, which is why this scores in the middle band rather than the top one. Its own measured case, a 70 percent lift in conversion rate and a 14.7 percent lift in average order value for Penn State Berkey Creamery, is explicitly Adobe Commerce.

Cost reasoning is honest but unquantified. The licence is described as typically starting in the low to mid five figures, the trap is named as comparing a licence fee to zero, and the verdict is that for most merchants below roughly mid seven figures of annual revenue with no serious B2B requirement the free edition wins the arithmetic. No figure is published for any self hosted cost line. The reverse path gets one sentence, that downgrades from Adobe Commerce to Open Source are technically possible but rare in practice.

Two things hold it back. Its Page Builder row reads not included, with third party equivalents existing, which Adobe's own documentation contradicts for every release from 2.4.3 onward, and that is a boundary error on the most consequential page it publishes. And it is not found in Hyvä's full agency register, while describing itself site wide as a certified Adobe partner with senior Magento engineers and more than 17 years on the platform, without stating a tier at a level.

elgentos does not publish an edition comparison, and on the heaviest criterion here it scores accordingly. What it publishes instead is the remedy, and for the buyer whose only real reason to consider Adobe Commerce is the B2B module it may be the most directly useful page in this ranking. Read on 28 September 2026, its B2B Suite page describes nineteen modular components built over ten years of B2B projects, built on top of the Magento core, open source where it can be, and explicitly without vendor lock in.

The components map closely onto what Adobe charges for. Company accounts joining multiple users to one organisation with automatic linking by email address or domain. Multi company accounts, where one customer account is linked to several companies and pricing and permissions follow whichever company is active, which is a requirement Adobe's own B2B documentation does not obviously cover. Company pricing with customer specific prices per company, tiered and quantity discounts, and contract prices held alongside standard Magento pricing rather than on top of it so that agreements, regular prices and promotions do not collide, usually fed from the customer's enterprise resource planning system. Company spending limits checked at checkout. Customer approval, so new accounts are vetted before they get access.

Its maintenance offer states the ownership trade off in the open, that freedom also means responsibility and a Magento platform has to be properly maintained through security updates, compatibility, performance and regression tests. The Total Care subscription covers all Magento core and extension updates, security patches and infrastructure upgrades monthly at a fixed price, deliberately separated from feature development so maintenance does not compete with revenue work, with a monthly report and changelog, automated test coverage, an annual re audit and advisory report, and every change traceable in the client's own repository. It publishes the four pillars the fixed monthly price is set from, shop size, existing code quality, the amount of customisation and integration, named as the most important pillar, and how well the agency knows the stack, with an annual review that can move the price down if the shop gets simpler. It publishes no figure.

Case evidence is named and unusually detailed but carries no measured outcome on the pages read. The ICTSpecialist case describes company pricing with monthly lease amounts shown on product pages, spending limits enforced in checkout, and several large customers each given a gated ordering portal inside a single Magento installation, with the frontend still on Luma and Hyvä on the roadmap. The Highlite case describes a new multi brand Hyvä frontend over an existing Magento 2 platform for the Infinity, Showtec and Wentex brands, with continuous integration, linting, image optimisation, real user Core Web Vitals monitoring and consent mode in place.

elgentos holds Platinum in Hyvä's full register and describes itself as a Magento B2B specialist since 2011. No Adobe tier at a level was found on the pages read, and every page read was in Dutch, which is worth knowing before a non Dutch speaking procurement team starts due diligence.

On a self hosted store the patch is the whole risk, and integer_net publishes the most detailed account of how one actually gets applied. Read on 28 September 2026, its patch article records deploying security patches to twenty shops in less than a day, covering all twenty active projects inside 24 hours, and states the standing rule for critical patches: all shops must be patched within 24 hours at most, with the team's record at around six hours.

The mechanism is published, not just the outcome. Criticality is decided by the Common Vulnerability Scoring System, with 9.0 to 10.0 treated as critical, read from Adobe's own security bulletin. Roles are assigned a month before each patch day, a patch day coordinator and a developer responsible for risk assessment, each with a named substitute in case of absence. Quality assurance runs from a per project test run template holding only high priority tests, so critical functionality can be cleared fast, with additional tests chosen from the developer's risk assessment of the changed code. The priority rule is written in capitals: a security update must not be blocked by any other release or by a Magento version upgrade, with no exceptions even when the regular release is already on staging.

It also publishes its own limit, which is what separates a process from a promise. Where there is doubt, the team does not wait for client acceptance testing, and the page says so and says why. It credits the whole approach to Adobe shipping security patches on a predictable two month cadence, which is a fair reading and one that a merchant planning their own maintenance window can act on directly.

Its Hyvä position is unusual and worth stating precisely. integer_net holds Platinum in Hyvä's full agency register, and its own partner page states that Hyvä's founder was part of the integer_net team when he had the idea for Hyvä, that the agency has followed the framework as a co creator since day one, and that it was among the first agencies to use Hyvä on a real project. It also states that whether a shop runs on Magento, on the Magento derived B2B product avanta, or on the community Magento distribution, the frontend is always Hyvä. It hosts the German language Hyvä community meetup and publishes a quick start Hyvä template of its own.

The gaps are large and they are the ones this page weights most. No Magento Open Source against Adobe Commerce comparison was found on the pages read, and no reverse migration path. No cost figure of any kind was found. Its reference page names real clients, among them a packaging manufacturer, a components brand, a public sector IT provider and a design retailer, and states that every store listed runs on Magento or on the Magento based avanta, but publishes no measured outcome for any of them. This is an agency that would score far higher on delivery than it does on disclosure.

Tigren's comparison, read on 28 September 2026, states the boundary without hedging: Magento Open Source is a solid platform out of the box, and some of the more advanced features found in Adobe Commerce are not included. It names the complete B2B suite with quotes and company accounts, Live Search powered by Adobe Sensei, seller assisted shopping, and built in reporting and analytics for which the free edition needs third party integrations. It does not address Page Builder or the version nuance around it on the page read, which is why it lands in the middle band rather than the upper one.

Its distinctive contribution is on cost, and it is genuinely useful. It publishes the licence range as 22,000 to 125,000 dollars a year varying with gross sales revenue, and then makes two points that no other agency in this set publishes. The first is that Adobe Commerce licence fees can often be negotiated, so two merchants may pay different amounts for the same product, which is a warning to treat any published tier as a reference rather than a quote. The second is the one small merchants get caught by: if revenue is relatively low the minimum licence fee still applies, and for a business not expecting rapid growth that floor may be prohibitive on its own. That is the sharpest single argument for the free edition anywhere on this page, and it comes from an agency that would earn more from the other answer.

It is also the only entry that names what community support actually consists of for a self hosted store, the Meet Magento events, the public Slack groups and the discussion forums, set against Adobe Commerce support arriving as a paid plan with service level agreements attached. That is an honest framing of the support boundary even though it stops short of any commitment of its own.

Patch ownership is where it thins out. Nothing was found on the pages read that states who applies patches to a Tigren maintained store, on what cadence, or within what window, so the criterion scores in the lowest band on the strength of the support contrast alone. No figure for hosting, extensions or a retainer was found either, so the cost score rests entirely on the licence band.

Case evidence could not be read. The case studies route on tigren.com served a WordPress sign in screen when it was requested on 28 September 2026, so nothing about Tigren's client work was verified and the criterion scores zero on that basis rather than on a finding that nothing is published. Tigren is listed at Bronze in Hyvä's full register, and no Adobe tier at a level was found on the pages read.

Strix publishes a short comparison that is more accurate than several longer ones. Read on 28 September 2026, it gets Page Builder right, stating that the tool works in both editions with drag and drop editing and live preview, that the open source version provides most of the basic content creation tools including layouts, media and text editing, and that Adobe Commerce adds dynamic content blocks, stronger analytics, integration with other Adobe products and personalisation and segmentation tooling. It gets Live Search right too, stating that Adobe Commerce offers it and that the free open source edition does not provide such functionality.

On B2B it takes a position most agencies avoid. Both editions can run a business to business shop, it says, and on the free edition the relevant extensions have to be installed and configured, which requires additional investment but lets small and medium companies get the capability without the licence. It then lists what Adobe Commerce builds in, custom pricing, catalogue and payment solutions, quote requests, purchase limits and credit limits, with user roles and permissions and buyer specific catalogues. That is a fair summary of the trade off rather than a push in one direction.

It is also plain about hosting, stating that Magento Open Source gives full freedom to choose a hosting provider but that hosting must be paid for separately because it is not part of the free package, and that Adobe Commerce brings automatic security patches and hosting inside the Adobe cloud. That is the boundary stated correctly in one sentence, and it is the whole of the cost reasoning published on the page read. No figure appears for hosting, extensions, a retainer or the licence, which is why the cost criterion scores near the floor.

Nothing was found on the pages read describing Strix's own patch cadence, response window or security operations for a self hosted store, so that criterion scores on the strength of the general statement alone. Case study routes on strix.net returned not found when requested on 28 September 2026, so no client work was verified and case evidence scores zero on that basis rather than on a finding that none exists.

Its partner standing is the best thing in its column after the comparison itself. Strix appears twice in Hyvä's full agency register, as Strix and as Strix NL, at Silver on both listings, which is the third highest of the register's five tiers. No Adobe tier at a level was found on the pages read.

Firebear Studio finishes last on total score and takes full marks on the criterion nobody else comes close on. It publishes the only step by step procedure in this set for moving a live store from Adobe Commerce down to Magento Open Source. Read on 28 September 2026, the article walks three routes: a free community project on GitHub that transfers entities between the editions and, as the page puts it, downgrades the database schema without altering the data; a Magento Stack Exchange thread on stripping the enterprise only files from an installation; and a longer third party guide with the branch creation and command sequence written out.

It also publishes the procedure's limit, which is the part that makes it usable rather than merely encouraging. The guide does not cover migrating custom themes and modules, and the page says so rather than leaving a reader to discover it mid migration. It names the trigger in commercial terms too, that a company which chose the paid path, is not using most of what it pays for, or needs to reduce platform cost, can move back down.

Reading the boundary backwards, as this article does, produces a clean list of what the licence buys: Page Builder, page hierarchy, company accounts with multiple users and custom structures, custom catalogues and pricing assigned to them, B2B quotes, quick order forms, requisition lists and reordering. Elsewhere on the same site company accounts are described flatly as a business to business entity available in the paid edition only. The licence is quoted as starting at 22,000 dollars a year for turnover under 1 million dollars. That is eight named capabilities and a price, in prose, which earns a middle band score on the boundary criterion without a comparison table anywhere.

Everything else is thin. Its services page describes ten years of work, more than two thousand customers, and a service list covering Magento 2 store building, search engine optimisation, security audits, migration from a long list of platforms, enterprise resource planning and customer relationship management integration and extension development, and it states that Firebear Studio is a Magento and Adobe partner without naming a tier. No patch cadence, response window or maintenance commitment was found on the pages read. No cost figure appears beyond the Adobe licence floor.

No named client case with a measured outcome was found on the pages read, and the portfolio route returned not found when requested on 28 September 2026. Firebear Studio does not appear in Hyvä's full agency register. Rank it last on disclosure and still read the downgrade article, because for the specific buyer who has already decided the licence is not earning its keep, it is the most actionable page any of these ten agencies publishes.